Table 15 Results comparing the RL agent with the Victim participant responses: Victim A (Aggregate of all stakeholder responses), and Victim B (Most Scenarios Completed), showing: Mean Tree Score (M.T.S), Mean ‘isTreeCorrectlyAnswered’ (M.C.A), Mean ‘isTreeWronglyAnswered’ (M.W.A), and Mean ‘isGatherAdditionalDataRequested’ (M.A.D).
Agent | M.T.S | M.C.A | M.W.A | M.A.D |
|---|---|---|---|---|
Victim A (collective) | − 0.8835 | 0.6580 | 0.3052 | 0.0967 |
Victim B (most scenarios) | − 1.3077 | 0.5922 | 0.3882 | 0.0000 |
RL agent | 1.4000 | 0.8800 | 0.1200 | 0.0000 |
